MXnet在Linux下安装小结

2015-11-06 技术学习

MXNet is a deep learning framework designed for both efficiency and flexibility. It allows you to mix the flavours of deep learning programs together to maximize the efficiency and your productivity.

mxnet是一个深度学习库,运算快,可扩展,又有R/Python/Julia/GO等完备的接口,是dmlc的cxxnet第二代的学习库,拥有以下特性:

多方位支持,各种程序+CPU+GPU

友好的设计(如果你用过xgboost)

多卡/多机均在支持范围内

有人说比Caffe依赖少(其实安装还是不太简单的)

在Linux,ubuntu下完成安装步骤,参考文档在这里:

Build

需要先进行包的构建(Python,Julia也需要先构建)

先安装依赖包:

支持c++11的编译包,比如 g++ >= 4.8

git - BLAS library.

opencv (可选)

sudo apt-get update
sudo apt-get install -y build-essential git libblas-dev libopencv-dev

然后Build

make -j4

R包安装

以下内容会自动安装相关的包(包含需要加载的和建议的)

Rscript -e "install.packages(‘devtools’, repo = ‘https://cran.rstudio.com')"
cd R-package
Rscript -e "library(devtools); library(methods); options(repos=c(CRAN=’https://cran.rstudio.com')); install_deps(dependencies = TRUE)"
cd ..
R CMD INSTALL mxnet_0.5.tar.gz

imager包有可能报错,使用以下的命令来安装fttw再安装

sudo apt-get install libfftw3-dev libX11-dev

之后,进入R,输入library(mxnet)不报错,就OK啦~

其他安装问题可以去这里提issue:https://github.com/dmlc/mxnet/issues